The e-m@diag module integrates research on the diagnosis and capitalization of knowledge achieved in the Department of Automatic Femto-st. It searches from a knowledge base of a diagnosis.

From the current state and past a system, it provides the most likely diagnosis and level of confidence in this diagnosis. The main features of this module are:

√ Knowledge extraction from data in a history of failure: data collection phase, filtering relevant variables and data mining;

√ Capitalization of knowledge through feedback, based on the tools of case-based reasoning (CBR)

√ Use of a model of knowledge for diagnosis, an approach based on type of case-based reasoning (based case raisonning – CBR)

√ Maintaining the case base or data.